基于粒子群优化算法的支持向量机研究

基于粒子群优化算法提出了一种通过优化支持向量机模型参数,建立更佳的支持向量机数学模型的方法. 针对双螺旋分类问题,分别利用基于粒子群优化算法所建立的支持向量机分类器和标准支持向量机分类器进行了仿真实验,利用所建立的评价体系对仿真实验所获得的实验数据进行了评估,评估结果表明基于粒子群优化算法的支持向量机分类器明显优于标准支持向量机分类器,其分类结果表明基于粒子群优化算法的支持向量机分类器提高了分类结果的准确性,同时也验证了基于粒子群优化算法的支持向量机分类器在数据分类中的有效性.      

融合GMM及SVM的特定音频事件高精度识别方法

针对特定音频事件识别中持续时间特别短的音频事件漏检概率高、识别速度较慢的问题,提出一种融合高斯混合模型（GMM）及支持向量机（SVM）的特定音频事件识别算法. 该方法利用GMM的统计分布描述能力和SVM的推广泛化能力,将GMM和SVM分别识别的结果进行融合处理,以手枪、步枪、机关枪等10类以上枪声为实验数据,无需针对每种枪声生成相应的识别模板,仅需训练生成2个识别模板. 实验结果表明,识别准确率达到92.71%. 该方法模板数量少,不需要多次训练,算法复杂度较低,不仅便于应用而且可大幅提升识别效率.      

镍基催化剂对乙烷氧化脱氢制乙烯性能的研究

用均匀沉淀法制备了片状复合氧化镍NiMO(M=Li,Sr,Y,Fe,La)催化剂,并研究了其对乙烷氧化脱氢制乙烯的催化性能.采用X射线衍射(XRD)、扫描电子显微镜(SEM)、程序升温还原(H2-TPR)和程序升温脱附(O2-TPD)等方法对镍基催化剂进行了表征.X射线衍射和扫描电子显微镜检测结果表明:助剂的加入对催化剂的形貌影响不大,仍为片状结构;平均粒径为5~10 nm.程序升温还原和脱附表征结果表明:助剂Li,Sr并未进入NiO晶格;而助剂Y,Fe,La与Ni之间具有相互作用,因而影响催化剂的理化性质和催化活性.其中,Fe掺杂的片状氧化镍催化剂对乙烯的选择性和收率均有所改善,340℃时NiFeO催化剂上的乙烯收率可达18.2%.     

光谱法研究EDTA-Fe(Ⅲ)与牛血清白蛋白的相互作用

应用紫外光谱法和荧光光谱法研究了乙二胺四乙酸铁(Ⅲ)(EDTA-Fe(Ⅲ))与牛血清白蛋白的相互作用.结果表明:EDTA-Fe(Ⅲ)可以进入牛血清白蛋白的疏水腔,使蛋白质非极性区的生色基暴露于极性溶剂;EDTA-Fe(Ⅲ)通过静态猝灭的方式使牛血清白蛋白的荧光强度显著降低;EDTA-Fe(Ⅲ)与牛血清白蛋白主要以氢键和范德华力相结合.测定了EDTA-Fe(Ⅲ)与牛血清白蛋白的结合常数KA和结合位点数n,利用Forster非辐射能量转移机制确定了牛血清白蛋白与EDTA-Fe(Ⅲ)的结合距离和能量转移效率,并使用同步荧光技术探讨了EDTA-Fe(Ⅲ)对牛血清白蛋白构象的影响.     

A machine learning-based method to detect fluorescent spots and an accelerated,parallel implementation of this method

Under most microscopes, the fluorescent sig- nals emitted from biological structures, such as vesicles, as well as the signals from single molecules will appear as small puncta, which contribute to a Gaussian-like distri- bution. Accurate segmentation of these spots will funda- mentally affect our interpretation of a specific biological progress. Because of the complicated backgrounds in images, many algorithms fail to identify all of the inter- esting signals; the tremendous amount of time required for algorithms to process large datasets can also decrease their utility. Here, we introduce an excellent robust detection method based on the machine learning algorithm Ada- Boost, which outperforms threshold-based segmentation, wavelets, and FDA under most situations. We also provide a GPU/multi-core CPU implementation of this algorithm; this implementation accelerates the algorithm approxi- mately 10- and 7-fold acceleration compared with a single CPU implementation. The great reduction of time should make this method a promising candidate in the processing of large datasets. Furthermore, we demonstrate the use of our algorithm on true fluorescent micrographs, and the results show that machine learning-based detection meth- ods outperform the four other previously reported methods.     

Finite element models and molecular dynamic simulations for studying the response of mast cell under mechanical activation

In micropipette aspiration experiment, increasing mechanical stress applied to cell membrane induced degranulation of mast cell as well as a current that could be inhibited by an inhibitor, which is specific for the transient receptor potential vanilloid （TRPVs） channels. To determine the sensitivity of TRPVs to membrane strain and tension, and to gain new insights into the activation mechanism of TRPVs, finite element models of mast cell and molecular dynamic simulations of human aquaporin-1 are presented. During the finite element simulations, the cell membrane sustained to micropipette aspiration was simulated, and the strain distribution along membrane thickness direction was obtained. Besides, combining the finite element models of osteoblast aspirated into micro- pipette and other compared models, we examined the relationship between cell mechanical stimulations and mechanical attributes and presented a new perspective to determine the cell equivalent elastic modulus. Consid- ering the indetermination of TRPV crystal structure, human aquaporin-1, one kind of the channel membrane proteins, substituting for TRPV, has been studied with molecular dynamic （MD） simulations, under different external lateral tensions which have been obtained in mast cell finite ele- ment simulations, to investigate the mechanical stimulation effects on the membrane channels. The simulations show that human aquaporin-1 undergoes significant conforma- tional change and expands in accordance with lateraltension, which not only confirms the tendency of the pre- vious electrophysiological experiments but also leads us to a better understanding of TRPVs. The multi-scale study combining finite element simulation and MD simulation is a significant breakthrough in the field of mechanical mechanism in cell system.     

Isolation and characterization of rl（t）, a gene that controls leaf rolling in rice

Moderate leaf rolling is one of the most important morphological traits in rice breeding for plant ideotype. Previous studies have shown that the rl（t） gene has a high breeding potential for developing hybrid-rice varieties with an ideal ideotype, because it leads to an appropriate leaf rolling index （LRI） of about 30 % in the heterozygous state, and had a positive effect on grain yield. In this study, we isolated rl（t） and performed a preliminary investigation of its function in regulating leaf rolling in rice. DNA sequencing identified a single base change （G to T） in the finely mapped region （11 kb） containing rl（t）, and this is located in 3′-untranslated region （3′-UTR） of the only predicted gene, Roc5 （Rice outermost cell-specific）. The expression level of Roc5 is significantly higher in the rl（t） mutant than in the wild-type. Using RNAi and overexpression analysis, we found that the expression level of Roc5 correlated with LRI and leaf bulliform area, and wasalso associated with leaf abaxial or adaxial rolling. These results confirmed that Roc5 controls leaf rolling in a dosagedependent manner. Bioinformatics analysis revealed a conserved 17-nt sequence （called the GU-rich element） in the 3′-UTR of HD-GL2 （Homeodomain-Glabra2） family genes including Roc5. Based on the model of this element in regulating mRNA stability in mammals, we speculate that the single nucleotide change in this element accounts for the higher expression level of Roc5 in the rl（t） mutant compared to the wild-type, which ultimately leads to adaxial rolling of the leaf. This discovery further enhances our knowledge of the molecular mechanisms underlying leaf rolling in rice.     

Characterization of cocrystals formed by grinding amino acids through terahertz time-domain spectroscopy

In this paper,we used terahertz time-domain spectroscopy(THz-TDS)over a range of 0.3–2.5 THz to investigate the formation of solid-state cocrystals of amino acids,formed by grinding mixtures of two different kinds of amino acids.For comparison,we prepared dual-layer samples,combined by pressing two single-component pellets together without grinding.In the ground-mixture samples,some extra absorption peaks appeared,different from the characteristic peaks of the pure components,but these peaks did not appear in the dual-layer samples.Thus,these extra absorption peaks in the THz range are unique features of cocrystals.From our results,we believe that THz-TDS is a promising technique to characterize cocrystals.     

4,4＇-二氯甲基联苯自聚合树脂合成条件优化及表征

通过正交实验对4,4＇-二氯甲基联苯（BCMBP）自聚合反应的工艺条件进行优化,对4种影响因素进行考察,得到影响产品比表面积大小因素的顺序为：预聚合时间、预聚合温度、单体浓度和催化剂用量.在优化条件下所得树脂的比表面积为1 512.66 m^2/g,树脂具有丰富的微孔和较高的聚合度.通过氮气吸附-脱附、SEM和热失重实验对树脂的孔径分布、微观形貌和热稳定性进行了研究.     

西瓜八氢番茄红素合成酶基因全长的克隆与表达分析

通过RT-PCR和RACE技术从西瓜果实中克隆八氢番茄红素合成酶（PSY）基因的cDNA全长,用生物信息学方法对其cDNA序列及推测氨基酸序列进行分析,并用实时荧光定量PCR技术研究PSY在不同瓤色西瓜果实发育过程中的表达情况.结果表明,PSY基因cDNA全长1 561 bp（GenBank登录号为KC166870）,其开放阅读框为1 266 bp,编码421个氨基酸.该基因及其推导的氨基酸序列与同为葫芦科的其他植物的PSY及氨基酸序列的同源性分别为87％和91％以上;序列分析显示,PSY氨基酸序列N末端存在转运肽信号序列.不同瓤色西瓜果实发育过程中PSY的表达存在明显差异,红瓤果实中的表达量最高,这表明PSY可能与红瓤果实中积累较多的类胡萝卜素有关;PSY的表达量均高于PSY-A,推测PSY基因主要负责果实中类胡萝卜素的合成.